K Means Clustering Interview Questions | Data Science Interview Questions On K means algorithm

[@jaysoni7812]

k refers to number of studies, in research world, that's why it use k in k means it means here we are studying on cluster, and n refers to number of outcomes in research world, so that's why in sklearn the parameter name is n_clusters because using that algorithm we wants outcome after running that algorithm it will gives us the n number of outcomes. I hope my research is correct 😊

[@UnfoldDataScience]

Three 👏👏👏 for you.

[@muhammedthayyib9202]

Oh great. I commented another answer, from common sense. 😀

[@SESHUNITR]

Very Good information from interview. keep doing thanks.

[@ramyaanand3668]

Exactly i was looking for same thing n i found it by u aman great video its has so much information....thnku so much aman keep exploring more

[@UnfoldDataScience]

Welcome Ramya.

[@vinodbiradar5266]

Would like to add one more point in KMEANS++, It internally analyzes the pattern of the data. Such as the spread of data (whether it is spherical, rectangle, oval etc.) and then initialize the centroids as explained.

[@UnfoldDataScience]

Yes Vinod. Thanks for adding it.

[@chandramouli5881]

This video helped me to understand K means. Thanks for the sharing

[@UnfoldDataScience]

Welcome Chandra.

[@Monuchaitu44]

Your videos were like cheat sheets for revising and remembering concepts very easily. Good and Great Job.

[@UnfoldDataScience]

Thanks Again. Please share in your data science groups if possible. That will be helpful for channel.

[@Monuchaitu44]

@@UnfoldDataScience Sure, I will make it to happen.

[@AnkitSingh-rd3he]

Due to its ubiquity, it is often called "the k-means algorithm"; it is also referred to as Lloyd's algorithm, particularly in the computer science community. It is sometimes also referred to as "naïve k-means", because there exist much faster alternatives

[@ajaykushwaha-je6mw]

very very informative video.

[@UnfoldDataScience]

Thanks Ajay.

[@kushalhu7189]

Brilliant Sir.....

[@UnfoldDataScience]

Thanks Kushal.

[@vishalbhapkar2359]

I have been following this channel since very beginning, now I can say this works pretty much for me, thanks @unfold data science and Mr. Aman Sir

[@UnfoldDataScience]

Thanks Vishal. :)

[@sudheeshe1384]

Thanks for the valuable contents

[@UnfoldDataScience]

Welcome Sudheesh :)

[@theethumnandrumpirartharav4137]

Awesome👍

[@UnfoldDataScience]

Thank you! Cheers!

[@sandipansarkar9211]

finished watching

[@himanshugautam1421]

Loved it.

[@UnfoldDataScience]

Thanks Himanshu :)

[@muhammedthayyib9202]

K stands for a number. That number in a whole number. It cannot have 1.5 number of cluster. In cross validation we use K-flod. Then why not n. n is like a random selection but K is like a choose the best number. Thank you aman

[@kaanchii123]

Thank you, you are a great teacher!

[@UnfoldDataScience]

You're very welcome!

[@nishanthvirat9044]

thank you so much sir

[@harithavalmiki9390]

Thank you so much for this explanation Aman!

[@UnfoldDataScience]

My pleasure

[@callmace]

Gr8

[@UnfoldDataScience]

Thanks Tausif :)

[@samruddhideshmukh5928]

Great video!!!

[@UnfoldDataScience]

Thanks Samruddhi.

[@souravbiswas6892]

Excellent video. I wish I would have seen this video before my final round of interview in Walmart. I became heartbroken when I was not selected :(

[@yash422vd]

N number of appreciation for your style of explanation is less, another great video. Your simplicity is your best asset.

[@UnfoldDataScience]

So nice of you Vishal. :)

[@vallimuthaiyah5098]

Thank you sir for such a valuable content and information on silhouette score.. please upload more interviews questions with hidden information.. K in k means clustering refers to number of clusters but not sure why it is called as using letter K

[@UnfoldDataScience]

Thanks a lot for watching.

[@pramodyadav4422]

Eagerly waiting to know why it's called K-Means

[@qazibasheer443]

The k-means clustering algorithm is called "k-means" because it specifically partitions the data into "k" clusters based on the mean of the data points. Other clustering algorithms may use different criteria for clustering, such as "n-means" which partitions the data into "n" clusters, or "s-means" which partitions the data based on the sum of squared distances. However, the k-means algorithm uses the mean of the data points to calculate the centroids, and it partitions the data into "k" clusters. Therefore, it is called k-means.

[@srprev]

Due to its ubiquity, it is often called "the k-means algorithm" :)

[@ArunSingh-bj6ux]

Hi , Could you cover the logic behind croston method forecasting

[@UnfoldDataScience]

Thanks Arun for feedback. Will add.

[@abithaanand7170]

Sir in 3:01 sec, I don't understand thw concept of how the convergence speed would be slow if two clusters are located near . Similarly, how would the convergence speed be faster if two clusters are not located together?

[@shashankhegde1258]

The elbow curve comes in the shape of K ?

[@praveenkuthuru7439]

In my opinion, the k-NN algorithm which was coined in 1951 tries to find out the nearest neighbor w.r.t. the distance function similar to k-Means which was coined post 1951, due to this reasons the 'k' is maintained as is since then and not any other letter. Is it right????

[@UnfoldDataScience]

This one i did not hear yet. What I know is, in statistics K is typically used for number of groups to analyze, hence.

[@DaughterOfGodJG]

K-Means: The name reflects the research perspective where K represents the grouping or study of clusters. n_clusters (Sklearn): This parameter name emphasizes the outcome-oriented view, where "N" is the number of cluster results to expect.

[@terryterry3733]

Super bro nice explanation and one thing i want to understand HOW KMEAN GETS OVERFIT? Pls give me the couple of details i didnt get the ans in internet .

[@UnfoldDataScience]

Overfitting is typically a problem in supervised learning, not k-means generally.

[@rishigupta2342]

Could you discuss interview question based on Decision tree & Random forest?

[@UnfoldDataScience]

Sure,

[@MohitGupta-sz4bh]

Very informative and helpful video Aman. keep up the good work. We would like to have this kind of interview questions and answers video on every Machine Learning Algorithm to crack the interview. Please do create video on other algorithms. Again superb a wonderful job :)

[@UnfoldDataScience]

Thanks Mohit. Sure.

[@bangarrajumuppidu8354]

who will take care of random picking points for initialization of centroid

[@UnfoldDataScience]

Python itself through "k-means" module

[@bangarrajumuppidu8354]

@@UnfoldDataScience thank u sir

[@sampathvinaykumarreddymajj790]

Need these kind of videos But why it is called K-Means ??

[@UnfoldDataScience]

Thanks Sampath. Pls do try to find out 😁😄

[@yt-1161]

In order to get people to confuse it with K nearest neighbors

[@amarmemane2583]

Hello sir, please make this kind of interview qun video on each machine learning algorithm if u want we are ready to fee for that also😊

[@UnfoldDataScience]

Thanks Amar for suggestion. Noted.

[@souravbiswas6892]

Because any parameter which can be tuned/tweaked, is represented by 'k' and not by a,b,c,d..

[@nishah4058]

Can u pls elaborate your answer?

[@abhinavkhandelwal1045]

I have a question.. if I have trained my data on 2 models for instance Random forest and logistic regression and it is giving me the same accuracy then what should be the basis to decide which one the two algorithms should I select for my data

[@UnfoldDataScience]

Depends on business need.

[@abhinavkhandelwal1045]

If a business sets free then what must be a parameter to strike out one of the Random forest and logistic regression if giving same accuracy?

[@jaysoni7812]

@@abhinavkhandelwal1045 choose whichever model is fast or give quick prediction. if your both model gives same accuracy then choose that model which is faster, it will help you to quick prediction

[@letslearndatasciencetogeth479]

Sir pls make a video on the mathematics behind silhouette score in detail

[@UnfoldDataScience]

I was thinking someone will ask, I will do it :)

[@letslearndatasciencetogeth479]

@@UnfoldDataScience thanks sir for the amazing explanation

[@vignesan4197]

Hello sir please any junior level data scientist job available please inform.